Hi guys,
Steve mentioned the Performance Golden Rule in his post, I’m wondering how to calculated the split across all 50,000 websites that HTTP Archive had crawled?
There is a TTFB column for table of latest_pages, is that the time for backend? Can you tell me that what’s the accurate difference between TTFB, renderStart, onLoad, fullyLoaded and visualComplete in HTTP Archive? I’m not sure about that.
Thanks in advance. 
In the page tables the TTFB is from the start of the page load until the first byte of the base page is returned. It will include any time for DNS, socket connections, SSL negotiation, redirects and server time.
To get JUST the backend time you’ll need to look at the timing for the TTFB for the base page in the requests table (TTFB at a request level is the server time). To be really accurate you’d also want to subtract the socket connect time to allow for the network RTT on the request which is included in the TTFB.
Hi all, apologies if this has been asked already, but is it possible to also get TTFB (and possibly other resource-level metrics) for the resources on a page? I don’t see it in the requests table, but perhaps I’m missing it.